web-centric tools

plotly provides a platform for online data analytics and visualizations; build on of HTML etc.; “grammar of graphics” in ggplot; convey a lot of info in a easy-to-understand format.

library(tidyverse)
library(plotly)

library(p8105.datasets)

Lets load the Airbnb datsets.

data("nyc_airbnb")

nyc_airbnb=nyc_airbnb|>
  mutate(
    rating=review_scores_location/2
  )|>
  select(
    neighbourhood_group, neighbourhood, rating, price, room_type, lat, long
  )|>
  filter(
    neighbourhood_group=="Manhattan",
    room_type=="Entire home/apt",
    price <= 500,
    price >=100
  )

nyc_airbnb
## # A tibble: 9,535 × 7
##    neighbourhood_group neighbourhood     rating price room_type        lat  long
##    <chr>               <chr>              <dbl> <dbl> <chr>          <dbl> <dbl>
##  1 Manhattan           Battery Park City   NA     400 Entire home/a… -74.0  40.7
##  2 Manhattan           Battery Park City   NA     225 Entire home/a… -74.0  40.7
##  3 Manhattan           Battery Park City   NA     220 Entire home/a… -74.0  40.7
##  4 Manhattan           Battery Park City    4.5   165 Entire home/a… -74.0  40.7
##  5 Manhattan           Battery Park City   NA     350 Entire home/a… -74.0  40.7
##  6 Manhattan           Battery Park City   NA     400 Entire home/a… -74.0  40.7
##  7 Manhattan           Battery Park City   NA     130 Entire home/a… -74.0  40.7
##  8 Manhattan           Battery Park City    5     225 Entire home/a… -74.0  40.7
##  9 Manhattan           Battery Park City   NA     245 Entire home/a… -74.0  40.7
## 10 Manhattan           Battery Park City   NA     150 Entire home/a… -74.0  40.7
## # ℹ 9,525 more rows

Let’s make a scatterplot but interactive at this time.

nyc_airbnb |>
  mutate(
    label=str_c("Price:&", price, "\nRating: ", rating)
  )|>
  plot_ly(
    x = ~lat, y=~long, color=~price, 
    text=~label,
    type = "scatter", mode = "markers", alpha=.5
  )

lets make a boxplot

nyc_airbnb |>
  mutate(
    neighbourhood=fct_reorder(neighbourhood,price)) |>

  plot_ly(
    x=~neighbourhood, y=~price,color=~neighbourhood,
    type="box", colors = "viridis"
  )

barcharts

nyc_airbnb |> 
  count(neighbourhood) |> 
  mutate(neighbourhood = fct_reorder(neighbourhood, n)) |> 
  plot_ly(x = ~neighbourhood, y = ~n, color = ~neighbourhood, type = "bar", colors = "viridis")

ggplotly

do not use this often!

ggp_scatterplot=
  nyc_airbnb|>
  ggplot(aes(x=lat, y=long, color=price))+
  geom_point()

ggplotly(ggp_scatterplot)